Abstract: How to obtain the compact and reasonable expression of all feasible disassembly paths has become the most important problem to be resolved in disassembly sequence planning. In order to optimize the sub-assembly clustering in planning of disassembly path, and to avoid combination explosion problem of calculation caused by traditional method，a theory of generalized axis conformed with the engineering semantic information and the concept about disassembly priority adjustment coefficient are put forward, considering the information about assembly structure and assembly process, quoting and extending the concept of matching relation which express the compact degree between parts. On operation procedure, it’s first to gain the coaxial membership matrix and its modified matrix by using the expression form of adjacency matrix, the coaxiality coefficient function and the disassembly priority distributing table. Then, it can be obtained considering related algorithms of fuzzy cluster analysis of transitivity fuzzy equivalent matrix and λ dynamic product clustering relationship graph. Using these, the sub-assembly clustering can be obtained in the end. The benefit of the proposed theory and method is illustrated through an example, which can improve the consistence with the actual engineering situation, and calculating work can be greatly reduced.

Key words: Disassembly sequence planning, Fuzzy clustering, Generalized axis theory, Sub-assembly